Analysis of Freesurfer parcellation Genetic Template, surface area and cortical thickness differences
library(easypackages)
libraries("here","heplots","ggplot2","ggseg","ggsegExtra","ggsegChen","tidyverse","patchwork","pheatmap","psych")
source(here("code","cohens_d.R"))
source(here("code","get_ggColorHue.R"))
datapath = here("data","tidy")
plotdir = here("plots")
pheno_data_gex =read.csv(file.path(datapath,"labelData_allGEXMRIsubs.csv"))
pheno_data_mri =read.csv(file.path(datapath,"labelData_all_MRI.csv"))
# compute total SA and CT mean
asd_mri = read.csv(file.path(datapath,"asd_mri_data.csv"))
td_mri = read.csv(file.path(datapath,"td_mri_data.csv"))
lh_sa_cols = 2:35
rh_sa_cols = 106:139
lh_ct_cols = 37:70
rh_ct_cols = 141:174
asd_mri$lh_SAtotal = rowSums(asd_mri[,lh_sa_cols])
asd_mri$rh_SAtotal = rowSums(asd_mri[,rh_sa_cols])
asd_mri$SAtotal = asd_mri$lh_SAtotal + asd_mri$rh_SAtotal
asd_mri$lh_CTmean = rowMeans(asd_mri[,lh_ct_cols])
asd_mri$rh_CTmean = rowMeans(asd_mri[,rh_ct_cols])
asd_mri$CTmean = rowMeans(asd_mri[,c(lh_ct_cols,rh_ct_cols)])
td_mri$lh_SAtotal = rowSums(td_mri[,lh_sa_cols])
td_mri$rh_SAtotal = rowSums(td_mri[,rh_sa_cols])
td_mri$SAtotal = td_mri$lh_SAtotal + td_mri$rh_SAtotal
td_mri$lh_CTmean = rowMeans(td_mri[,lh_ct_cols])
td_mri$rh_CTmean = rowMeans(td_mri[,rh_ct_cols])
td_mri$CTmean = rowMeans(td_mri[,c(lh_ct_cols,rh_ct_cols)])
all_mri = rbind(td_mri,asd_mri)
all_mri$subjectId = factor(all_mri$X)
fs_data = read.csv(file.path(datapath,"FSparc_GenTemp_all.csv"))
fs_data$Group = factor(fs_data$Group)
# region_names = colnames(fs_data)[5:ncol(fs_data)]
# region_names = c("CortexVol","SAtotal","CTmean",region_names)
region_names = c("CortexVol","SAtotal","CTmean")
fs_data = merge(fs_data, pheno_data_mri[,c("subjectId","sex","scan_age")], by="subjectId")
fs_data = merge(fs_data, all_mri[,c("subjectId","SAtotal","CTmean")], by="subjectId")
#------------------------------------------------------------------------------
# Remove the longitudinal duplicates
dup_df = data.frame(t(table(fs_data$subjectId)))
colnames(dup_df)[2:3] = c("subjectId","n")
dup_subs = dup_df$subjectId[dup_df$n>1]
rows2keep = numeric()
rows2remove = numeric()
for (isub in dup_subs){
dup_indx = which(fs_data$subjectId==isub)
rows2keep = c(rows2keep,dup_indx[1])
rows2remove = c(rows2remove,dup_indx[2:length(dup_indx)])
}
fs_data = fs_data[-rows2remove,]
#------------------------------------------------------------------------------
new_fs_data = fs_data
colnames2use = c("Fstat","pval","etasq","fdrq","es.TD_vs_Good","es.TD_vs_Poor","es.Good_vs_Poor",
"tstat.TD_vs_Good","tstat.TD_vs_Poor","tstat.Good_vs_Poor",
"pval.TD_vs_Good","pval.TD_vs_Poor","pval.Good_vs_Poor")
output_res = data.frame(matrix(nrow=length(region_names), ncol=length(colnames2use)))
rownames(output_res) = region_names
colnames(output_res) = colnames2use
for (icol in 1:length(region_names)){
y_var = region_names[icol]
if(is.element(y_var,c("CortexVol","SAtotal","CTmean"))){
# make formula
form2use = as.formula(sprintf("%s ~ Group + sex + scan_age", y_var))
full_model = model.matrix(~0+ Group + sex + scan_age, data = fs_data)
covname2use = c("sexMale","scan_age")
}else{
# make formula
form2use = as.formula(sprintf("%s ~ Group + CortexVol + sex + scan_age", y_var))
full_model = model.matrix(~0+ Group + CortexVol + sex + scan_age, data = fs_data)
covname2use = c("CortexVol","sexMale","scan_age")
# #--------------------------------------------------------------------------
# # remove CortexVol as covariate since GCLUST is already globally controlled for
# # make formula
# form2use = as.formula(sprintf("%s ~ Group + sex + scan_age", y_var))
# full_model = model.matrix(~0+ Group + sex + scan_age, data = fs_data)
# covname2use = c("sexMale","scan_age")
# #--------------------------------------------------------------------------
}
# run linear model
mod2use = lm(formula = form2use, data = fs_data)
# run an ANOVA
res2use = anova(mod2use)
# grab ANOVA results and store into output_res
output_res[y_var, "Fstat"] = res2use["Group", "F value"]
output_res[y_var, "pval"] = res2use["Group", "Pr(>F)"]
# compute partial eta-squared as the effect size of the interaction effect
eta_sq_res = etasq(mod2use)
output_res[icol, "etasq"] = eta_sq_res["Group","Partial eta^2"]
# remove covariates before effect size computation
beta1 = mod2use$coefficients[covname2use, drop = FALSE]
beta1[is.na(beta1)] = 0
new_fs_data[,y_var] = as.numeric(t(fs_data[,y_var] - beta1 %*% t(full_model[,covname2use])))
# compute effect sizes
mask1 = fs_data$Group=="Good"
mask2 = fs_data$Group=="TD"
output_res[y_var, "es.TD_vs_Good"] = cohens_d(new_fs_data[mask2,y_var], new_fs_data[mask1,y_var])
d2use = subset(fs_data, mask1 | mask2)
mod2use = lm(formula = form2use, data = d2use)
res = summary(mod2use)
output_res[y_var, "tstat.TD_vs_Good"] = res$coefficients[2,"t value"]
output_res[y_var, "pval.TD_vs_Good"] = res$coefficients[2,"Pr(>|t|)"]
mask1 = fs_data$Group=="Poor"
mask2 = fs_data$Group=="TD"
output_res[y_var, "es.TD_vs_Poor"] = cohens_d(new_fs_data[mask2,y_var], new_fs_data[mask1,y_var])
d2use = subset(fs_data, mask1 | mask2)
mod2use = lm(formula = form2use, data = d2use)
res = summary(mod2use)
output_res[y_var, "tstat.TD_vs_Poor"] = res$coefficients[2,"t value"]
output_res[y_var, "pval.TD_vs_Poor"] = res$coefficients[2,"Pr(>|t|)"]
mask1 = fs_data$Group=="Good"
mask2 = fs_data$Group=="Poor"
output_res[y_var, "es.Good_vs_Poor"] = cohens_d(new_fs_data[mask1,y_var], new_fs_data[mask2,y_var])
d2use = subset(fs_data, mask1 | mask2)
mod2use = lm(formula = form2use, data = d2use)
res = summary(mod2use)
output_res[y_var, "tstat.Good_vs_Poor"] = res$coefficients[2,"t value"]
output_res[y_var, "pval.Good_vs_Poor"] = res$coefficients[2,"Pr(>|t|)"]
}
output_res$fdrq = p.adjust(output_res$pval, method = "fdr")
output_res$fdrq.TD_vs_Good = p.adjust(output_res$pval.TD_vs_Good, method = "fdr")
output_res$fdrq.TD_vs_Poor = p.adjust(output_res$pval.TD_vs_Poor, method = "fdr")
output_res$fdrq.Good_vs_Poor = p.adjust(output_res$pval.Good_vs_Poor, method = "fdr")
output_res[order(output_res$pval),]
## Fstat pval etasq fdrq es.TD_vs_Good
## SAtotal 15.397759 1.143503e-06 0.072173002 3.430509e-06 -0.27132289
## CortexVol 14.303102 2.741889e-06 0.075345308 4.112833e-06 -0.26212136
## CTmean 2.803486 6.464027e-02 0.002252612 6.464027e-02 0.07724623
## es.TD_vs_Poor es.Good_vs_Poor tstat.TD_vs_Good tstat.TD_vs_Poor
## SAtotal -0.7113176 -0.40788261 -1.2806955 -2.8440667
## CortexVol -0.7087278 -0.42651049 -1.1088191 -2.9849992
## CTmean 0.1155233 0.03662885 0.3724289 0.1765433
## tstat.Good_vs_Poor pval.TD_vs_Good pval.TD_vs_Poor pval.Good_vs_Poor
## SAtotal 1.7936785 0.2039557 0.005637100 0.07706336
## CortexVol 1.8842593 0.2707896 0.003748508 0.06356704
## CTmean -0.2434033 0.7105462 0.860307799 0.80838511
## fdrq.TD_vs_Good fdrq.TD_vs_Poor fdrq.Good_vs_Poor
## SAtotal 0.4061844 0.008455649 0.1155950
## CortexVol 0.4061844 0.008455649 0.1155950
## CTmean 0.7105462 0.860307799 0.8083851
Make plots of total cortical volume, surface area, and mean cortical thickness
dotSize = 6
fontSize=28
fdr_thresh = 0.05
output_res_fdr = output_res[output_res$fdrq<=fdr_thresh,]
output_res_fdr[order(output_res_fdr$pval),]
## Fstat pval etasq fdrq es.TD_vs_Good
## SAtotal 15.39776 1.143503e-06 0.07217300 3.430509e-06 -0.2713229
## CortexVol 14.30310 2.741889e-06 0.07534531 4.112833e-06 -0.2621214
## es.TD_vs_Poor es.Good_vs_Poor tstat.TD_vs_Good tstat.TD_vs_Poor
## SAtotal -0.7113176 -0.4078826 -1.280695 -2.844067
## CortexVol -0.7087278 -0.4265105 -1.108819 -2.984999
## tstat.Good_vs_Poor pval.TD_vs_Good pval.TD_vs_Poor pval.Good_vs_Poor
## SAtotal 1.793679 0.2039557 0.005637100 0.07706336
## CortexVol 1.884259 0.2707896 0.003748508 0.06356704
## fdrq.TD_vs_Good fdrq.TD_vs_Poor fdrq.Good_vs_Poor
## SAtotal 0.4061844 0.008455649 0.115595
## CortexVol 0.4061844 0.008455649 0.115595
# Plot Total Cortical Volume
cols2use = c("#3d3d3d","#828282","#b3b3b3")
region = "CortexVol"
reg_name = "Total Cortical Volume"
metric_name = "Volume"
ylims = c(450000,700000)
p1 = ggplot(data = fs_data, aes_string(x = "Group", y = region, fill = "Group"))
p1 = p1 + geom_jitter(size = dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
scale_fill_manual(values=cols2use) + guides(colour = FALSE, fill = FALSE) +
xlab("Group") + ylab(metric_name) + ggtitle(reg_name) + ylim(ylims) +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = here("plots","TotalCorticalVolume.pdf"))
p1
# Plot Total Surface Area
cols2use = c("#3d3d3d","#828282","#b3b3b3")
region = "SAtotal"
reg_name = "Total Surface Area"
metric_name = "SA"
ylims = c(90000,220000)
p1 = ggplot(data = fs_data, aes_string(x = "Group", y = region, fill = "Group"))
p1 = p1 + geom_jitter(size = dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
scale_fill_manual(values=cols2use) + guides(colour = FALSE, fill = FALSE) +
xlab("Group") + ylab(metric_name) + ggtitle(reg_name) + ylim(ylims) +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = here("plots","TotalSurfaceArea.pdf"))
p1
# Plot Mean Cortical Thickness
cols2use = c("#3d3d3d","#828282","#b3b3b3")
region = "CTmean"
reg_name = "Mean Cortical Thickness"
metric_name = "CT"
ylims = c(2.75,3.75)
p1 = ggplot(data = fs_data, aes_string(x = "Group", y = region, fill = "Group"))
p1 = p1 + geom_jitter(size = dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
scale_fill_manual(values=cols2use) + guides(colour = FALSE, fill = FALSE) +
xlab("Group") + ylab(metric_name) + ggtitle(reg_name) + ylim(ylims) +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = here("plots","MeanCorticalThickness.pdf"),width=7,height=5)
p1
Do modeling of local differences without adjustment for total cortical volume because GCLUST is already globally corrected.
# compute total SA and CT mean
asd_mri = read.csv(file.path(datapath,"asd_mri_data.csv"))
td_mri = read.csv(file.path(datapath,"td_mri_data.csv"))
lh_sa_cols = 2:35
rh_sa_cols = 106:139
lh_ct_cols = 37:70
rh_ct_cols = 141:174
asd_mri$lh_SAtotal = rowSums(asd_mri[,lh_sa_cols])
asd_mri$rh_SAtotal = rowSums(asd_mri[,rh_sa_cols])
asd_mri$SAtotal = asd_mri$lh_SAtotal + asd_mri$rh_SAtotal
asd_mri$lh_CTmean = rowMeans(asd_mri[,lh_ct_cols])
asd_mri$rh_CTmean = rowMeans(asd_mri[,rh_ct_cols])
asd_mri$CTmean = rowMeans(asd_mri[,c(lh_ct_cols,rh_ct_cols)])
td_mri$lh_SAtotal = rowSums(td_mri[,lh_sa_cols])
td_mri$rh_SAtotal = rowSums(td_mri[,rh_sa_cols])
td_mri$SAtotal = td_mri$lh_SAtotal + td_mri$rh_SAtotal
td_mri$lh_CTmean = rowMeans(td_mri[,lh_ct_cols])
td_mri$rh_CTmean = rowMeans(td_mri[,rh_ct_cols])
td_mri$CTmean = rowMeans(td_mri[,c(lh_ct_cols,rh_ct_cols)])
all_mri = rbind(td_mri,asd_mri)
all_mri$subjectId = factor(all_mri$X)
fs_data = read.csv(file.path(datapath,"FSparc_GenTemp_all.csv"))
fs_data$Group = factor(fs_data$Group)
region_names = colnames(fs_data)[5:ncol(fs_data)]
# region_names = c("CortexVol","SAtotal","CTmean",region_names)
fs_data = merge(fs_data, pheno_data_mri[,c("subjectId","sex","scan_age")], by="subjectId")
fs_data = merge(fs_data, all_mri[,c("subjectId","SAtotal","CTmean")], by="subjectId")
#------------------------------------------------------------------------------
# Remove duplicates
dup_df = data.frame(t(table(fs_data$subjectId)))
colnames(dup_df)[2:3] = c("subjectId","n")
dup_subs = dup_df$subjectId[dup_df$n>1]
rows2keep = numeric()
rows2remove = numeric()
for (isub in dup_subs){
dup_indx = which(fs_data$subjectId==isub)
rows2keep = c(rows2keep,dup_indx[1])
rows2remove = c(rows2remove,dup_indx[2:length(dup_indx)])
}
fs_data = fs_data[-rows2remove,]
#------------------------------------------------------------------------------
new_fs_data = fs_data
colnames2use = c("Fstat","pval","etasq","fdrq","es.TD_vs_Good","es.TD_vs_Poor","es.Good_vs_Poor",
"tstat.TD_vs_Good","tstat.TD_vs_Poor","tstat.Good_vs_Poor",
"pval.TD_vs_Good","pval.TD_vs_Poor","pval.Good_vs_Poor")
output_res = data.frame(matrix(nrow=length(region_names), ncol=length(colnames2use)))
rownames(output_res) = region_names
colnames(output_res) = colnames2use
for (icol in 1:length(region_names)){
y_var = region_names[icol]
# make formula
form2use = as.formula(sprintf("%s ~ Group + sex + scan_age", y_var))
full_model = model.matrix(~0+ Group + sex + scan_age, data = fs_data)
covname2use = c("sexMale","scan_age")
# run linear model
mod2use = lm(formula = form2use, data = fs_data)
# run an ANOVA
res2use = anova(mod2use)
# grab ANOVA results and store into output_res
output_res[y_var, "Fstat"] = res2use["Group", "F value"]
output_res[y_var, "pval"] = res2use["Group", "Pr(>F)"]
# compute partial eta-squared as the effect size of the interaction effect
eta_sq_res = etasq(mod2use)
output_res[icol, "etasq"] = eta_sq_res["Group","Partial eta^2"]
# remove covariates before effect size computation
beta1 = mod2use$coefficients[covname2use, drop = FALSE]
beta1[is.na(beta1)] = 0
new_fs_data[,y_var] = as.numeric(t(fs_data[,y_var] - beta1 %*% t(full_model[,covname2use])))
# compute effect sizes
mask1 = fs_data$Group=="Good"
mask2 = fs_data$Group=="TD"
output_res[y_var, "es.TD_vs_Good"] = cohens_d(new_fs_data[mask2,y_var], new_fs_data[mask1,y_var])
d2use = subset(fs_data, mask1 | mask2)
mod2use = lm(formula = form2use, data = d2use)
res = summary(mod2use)
output_res[y_var, "tstat.TD_vs_Good"] = res$coefficients[2,"t value"]
output_res[y_var, "pval.TD_vs_Good"] = res$coefficients[2,"Pr(>|t|)"]
mask1 = fs_data$Group=="Poor"
mask2 = fs_data$Group=="TD"
output_res[y_var, "es.TD_vs_Poor"] = cohens_d(new_fs_data[mask2,y_var], new_fs_data[mask1,y_var])
d2use = subset(fs_data, mask1 | mask2)
mod2use = lm(formula = form2use, data = d2use)
res = summary(mod2use)
output_res[y_var, "tstat.TD_vs_Poor"] = res$coefficients[2,"t value"]
output_res[y_var, "pval.TD_vs_Poor"] = res$coefficients[2,"Pr(>|t|)"]
mask1 = fs_data$Group=="Good"
mask2 = fs_data$Group=="Poor"
output_res[y_var, "es.Good_vs_Poor"] = cohens_d(new_fs_data[mask1,y_var], new_fs_data[mask2,y_var])
d2use = subset(fs_data, mask1 | mask2)
mod2use = lm(formula = form2use, data = d2use)
res = summary(mod2use)
output_res[y_var, "tstat.Good_vs_Poor"] = res$coefficients[2,"t value"]
output_res[y_var, "pval.Good_vs_Poor"] = res$coefficients[2,"Pr(>|t|)"]
}
output_res$fdrq = p.adjust(output_res$pval, method = "fdr")
output_res$fdrq.TD_vs_Good = p.adjust(output_res$pval.TD_vs_Good, method = "fdr")
output_res$fdrq.TD_vs_Poor = p.adjust(output_res$pval.TD_vs_Poor, method = "fdr")
output_res$fdrq.Good_vs_Poor = p.adjust(output_res$pval.Good_vs_Poor, method = "fdr")
write.csv(output_res, file = here("results","GCLUST_regional_group_differences_stats.csv"))
output_res[order(output_res$pval),]
## Fstat pval etasq
## middletemporal.rh.thickness 5.94103928 0.003480438 0.0734094108
## posterolateraltemporal.rh.area 5.46548677 0.005369823 0.0598876648
## anteromedialtemporal.lh.area 5.15535591 0.007137177 0.0658046617
## superiorparietal.rh.area 4.07830238 0.019379782 0.0287779414
## medialprefrontal.lh.thickness 4.06769013 0.019573122 0.0275112055
## middletemporal.lh.thickness 3.41762916 0.036069477 0.0343140556
## superiorparietal.lh.area 3.31498536 0.039747640 0.0330204534
## dorsomedialfrontal.lh.area 3.15695112 0.046171733 0.0404887240
## medialprefrontal.rh.thickness 2.92870020 0.057364440 0.0129543380
## temporalpole.rh.thickness 2.76347611 0.067157819 0.0229000518
## orbitofrontal.rh.area 2.56863271 0.080920331 0.0853698327
## temporalpole.lh.thickness 2.38151980 0.096838507 0.0370636668
## occipital.lh.area 2.24561269 0.110368376 0.0271548766
## posterolateraltemporal.lh.area 1.93258720 0.149327978 0.0140083037
## inferiorparietal.lh.area 1.89942858 0.154201588 0.0328601053
## dorsolateralprefrotal.lh.thickness 1.86898803 0.158818058 0.0112166280
## inferiorparietal.lh.thickness 1.84940380 0.161862175 0.0293629309
## orbitofrontal.lh.area 1.77668475 0.173694196 0.0542587241
## perisylvian.lh.thickness 1.76919001 0.174962619 0.0083808211
## superiortemporal.rh.area 1.70160553 0.186834976 0.0080576156
## dorsolateralprefrotal.rh.thickness 1.69970989 0.187179541 0.0113713568
## anteromedialtemporal.rh.area 1.68188886 0.190450595 0.0078339922
## dorsomedialfrontal.rh.area 1.47819586 0.232240462 0.0313502126
## inferiorparietal.rh.area 1.46476311 0.235304210 0.0322111775
## motor_premotor_SMA.rh.thickness 1.30638497 0.274684555 0.0447032529
## motor_premotor_SMA.lh.thickness 1.18615421 0.309007959 0.0610710194
## perisylvian.rh.thickness 1.15784633 0.317705086 0.0132777649
## ventralfrontal.rh.thickness 1.12958718 0.326635646 0.0037612076
## ventromedialoccipital.lh.thickness 1.03790036 0.357408138 0.0338876279
## inferiorparietal.rh.thickness 1.03713528 0.357676959 0.0196799443
## ventromedialoccipital.rh.thickness 0.87775545 0.418411359 0.0162415410
## superiorparietal.rh.thickness 0.63051356 0.534102064 0.0105049232
## medialtemporal.lh.thickness 0.59903257 0.551004680 0.0101886096
## dorsolateralprefrontal.lh.area 0.54820187 0.579453005 0.0117343658
## motorpremotor.lh.area 0.50722783 0.603469423 0.0069401882
## superiorparietal.lh.thickness 0.47653492 0.622120938 0.0080707604
## dorsolateralprefrontal.rh.area 0.43942349 0.645457966 0.0040848178
## occipital.rh.area 0.37917411 0.685257301 0.0066204803
## occipital.rh.thickness 0.32871018 0.720508629 0.0041295970
## motorpremotor.rh.area 0.32742064 0.721433208 0.0105939533
## ventralfrontal.lh.thickness 0.30587253 0.737062641 0.0030880881
## precuneus.rh.area 0.28560197 0.752079731 0.0035408362
## occipital.lh.thickness 0.28391299 0.753344940 0.0054019402
## medialtemporal.rh.thickness 0.21417637 0.807519105 0.0003133628
## parsopercularis.lh.area 0.15301539 0.858286507 0.0053473121
## parsopercularis.rh.area 0.15014302 0.860749032 0.0008351415
## precuneus.lh.area 0.08999012 0.914002879 0.0017483701
## superiortemporal.lh.area 0.01223912 0.987836730 0.0029551953
## fdrq es.TD_vs_Good es.TD_vs_Poor
## middletemporal.rh.thickness 0.1141948 -0.663418297 -0.392605943
## posterolateraltemporal.rh.area 0.1141948 0.030295479 -0.535892781
## anteromedialtemporal.lh.area 0.1141948 0.105637215 -0.533075090
## superiorparietal.rh.area 0.1879020 -0.101374144 0.311313007
## medialprefrontal.lh.thickness 0.1879020 0.118687302 0.420863328
## middletemporal.lh.thickness 0.2725552 -0.149600918 -0.519216132
## superiorparietal.lh.area 0.2725552 -0.141813175 0.330575141
## dorsomedialfrontal.lh.area 0.2770304 -0.360530708 0.121131105
## medialprefrontal.rh.thickness 0.3059437 0.108139104 0.282925991
## temporalpole.rh.thickness 0.3223575 -0.294484547 -0.368873339
## orbitofrontal.rh.area 0.3531069 0.669848517 0.631671827
## temporalpole.lh.thickness 0.3873540 0.022831450 -0.431998283
## occipital.lh.area 0.4075140 0.417678877 0.183761925
## posterolateraltemporal.lh.area 0.4155286 -0.108519406 -0.306559251
## inferiorparietal.lh.area 0.4155286 -0.215460672 -0.459224327
## dorsolateralprefrotal.lh.thickness 0.4155286 0.119445339 0.275104346
## inferiorparietal.lh.thickness 0.4155286 0.188388475 -0.258606627
## orbitofrontal.lh.area 0.4155286 0.214206795 0.585759310
## perisylvian.lh.thickness 0.4155286 -0.210011882 -0.170481439
## superiortemporal.rh.area 0.4155286 0.209040493 0.153640728
## dorsolateralprefrotal.rh.thickness 0.4155286 0.215372099 -0.006725081
## anteromedialtemporal.rh.area 0.4155286 -0.096075705 -0.232557068
## dorsomedialfrontal.rh.area 0.4706084 0.060393743 0.482914856
## inferiorparietal.rh.area 0.4706084 -0.464685948 -0.104099702
## motor_premotor_SMA.rh.thickness 0.5273943 0.319915972 0.547023221
## motor_premotor_SMA.lh.thickness 0.5599468 0.448981145 0.648166073
## perisylvian.rh.thickness 0.5599468 0.078911716 -0.206641773
## ventralfrontal.rh.thickness 0.5599468 -0.148932409 -0.063507467
## ventromedialoccipital.lh.thickness 0.5722831 -0.304107726 -0.503207370
## inferiorparietal.rh.thickness 0.5722831 0.330777772 0.055870712
## ventromedialoccipital.rh.thickness 0.6478627 -0.293894577 -0.261454543
## superiorparietal.rh.thickness 0.8011531 0.219385212 0.012428666
## medialtemporal.lh.thickness 0.8014614 0.044354354 0.241938279
## dorsolateralprefrontal.lh.area 0.8180513 0.122140978 0.270635677
## motorpremotor.lh.area 0.8276152 -0.032713682 0.165653399
## superiorparietal.lh.thickness 0.8294946 -0.100464989 0.138091765
## dorsolateralprefrontal.rh.area 0.8373509 0.005594604 -0.132562680
## occipital.rh.area 0.8409432 0.157964681 0.183810743
## occipital.rh.thickness 0.8409432 -0.012097343 0.140113306
## motorpremotor.rh.area 0.8409432 -0.232932678 -0.188478535
## ventralfrontal.lh.thickness 0.8409432 -0.135651383 -0.111708180
## precuneus.rh.area 0.8409432 -0.015804628 0.118539861
## occipital.lh.thickness 0.8409432 -0.061960608 0.118584159
## medialtemporal.rh.thickness 0.8809299 -0.033634161 0.006069805
## parsopercularis.lh.area 0.8981729 -0.049461374 -0.206013379
## parsopercularis.rh.area 0.8981729 -0.057073500 0.001465130
## precuneus.lh.area 0.9334497 -0.026753735 0.075252332
## superiortemporal.lh.area 0.9878367 -0.087919189 -0.137690521
## es.Good_vs_Poor tstat.TD_vs_Good
## middletemporal.rh.thickness 0.308293704 -2.85110017
## posterolateraltemporal.rh.area -0.578778313 0.36311947
## anteromedialtemporal.lh.area -0.627195070 0.37894890
## superiorparietal.rh.area 0.438346980 -0.63925964
## medialprefrontal.lh.thickness 0.274891792 0.41679478
## middletemporal.lh.thickness -0.296831346 -0.66119236
## superiorparietal.lh.area 0.440699732 -0.67574253
## dorsomedialfrontal.lh.area 0.480764273 -1.63798702
## medialprefrontal.rh.thickness 0.175740176 0.64582311
## temporalpole.rh.thickness -0.057108681 -1.34457649
## orbitofrontal.rh.area -0.047911120 2.95765605
## temporalpole.lh.thickness -0.430949000 0.05637093
## occipital.lh.area -0.221406959 1.73969339
## posterolateraltemporal.lh.area -0.187309136 -0.49040595
## inferiorparietal.lh.area -0.242790899 -0.91920137
## dorsolateralprefrotal.lh.thickness 0.144043968 0.39221574
## inferiorparietal.lh.thickness -0.426697227 1.03758906
## orbitofrontal.lh.area 0.385080584 0.92711527
## perisylvian.lh.thickness 0.035500517 -0.94804045
## superiortemporal.rh.area -0.056252035 1.09667446
## dorsolateralprefrotal.rh.thickness -0.245041864 0.62950223
## anteromedialtemporal.rh.area -0.121524702 -0.54955871
## dorsomedialfrontal.rh.area 0.330214373 0.25946827
## inferiorparietal.rh.area 0.298412101 -2.24081659
## motor_premotor_SMA.rh.thickness 0.215161604 1.52075062
## motor_premotor_SMA.lh.thickness 0.169031261 2.12900175
## perisylvian.rh.thickness -0.291323899 0.25516795
## ventralfrontal.rh.thickness 0.086835099 -0.89476216
## ventromedialoccipital.lh.thickness -0.122955125 -1.32562092
## inferiorparietal.rh.thickness -0.279450323 1.52372353
## ventromedialoccipital.rh.thickness 0.046647076 -1.22983796
## superiorparietal.rh.thickness -0.230412846 1.11877053
## medialtemporal.lh.thickness 0.201395580 0.22010597
## dorsolateralprefrontal.lh.area 0.151199701 0.67279733
## motorpremotor.lh.area 0.193878295 -0.20536507
## superiorparietal.lh.thickness 0.209895159 -0.28637751
## dorsolateralprefrontal.rh.area -0.150628075 0.09706286
## occipital.rh.area 0.046920394 0.56402834
## occipital.rh.thickness 0.144938773 -0.04980532
## motorpremotor.rh.area 0.049542202 -0.94441590
## ventralfrontal.lh.thickness 0.008053005 -0.82119741
## precuneus.rh.area 0.158720852 -0.11229016
## occipital.lh.thickness 0.191400339 -0.37207575
## medialtemporal.rh.thickness 0.039101046 -0.03002987
## parsopercularis.lh.area -0.123165860 -0.29915811
## parsopercularis.rh.area 0.066652372 -0.21632630
## precuneus.lh.area 0.104640544 -0.07671375
## superiortemporal.lh.area -0.046795143 -0.06771132
## tstat.TD_vs_Poor tstat.Good_vs_Poor
## middletemporal.rh.thickness -1.55784151 -1.27463943
## posterolateraltemporal.rh.area -2.18864474 2.70658431
## anteromedialtemporal.lh.area -1.89885775 2.70178261
## superiorparietal.rh.area 1.52415173 -1.92100876
## medialprefrontal.lh.thickness 1.79227030 -1.20242122
## middletemporal.lh.thickness -1.94016360 1.36426487
## superiorparietal.lh.area 1.51281550 -1.83973682
## dorsomedialfrontal.lh.area 0.41544508 -2.20890355
## medialprefrontal.rh.thickness 1.10578637 -0.69398289
## temporalpole.rh.thickness -1.35147102 0.46949803
## orbitofrontal.rh.area 2.90748713 0.25800500
## temporalpole.lh.thickness -1.76002661 1.96294321
## occipital.lh.area 0.52774695 0.85917149
## posterolateraltemporal.lh.area -0.95004682 0.94694197
## inferiorparietal.lh.area -1.96219360 1.06956480
## dorsolateralprefrotal.lh.thickness 1.35489026 -0.59374712
## inferiorparietal.lh.thickness -1.18684886 1.84689550
## orbitofrontal.lh.area 2.43578917 -1.69232724
## perisylvian.lh.thickness -0.86889078 -0.18510829
## superiortemporal.rh.area 0.32997478 0.25054848
## dorsolateralprefrotal.rh.thickness 0.14948667 0.95760379
## anteromedialtemporal.rh.area -0.86457779 0.49201307
## dorsomedialfrontal.rh.area 2.12769481 -1.38047753
## inferiorparietal.rh.area -0.16728797 -1.24213808
## motor_premotor_SMA.rh.thickness 2.23692644 -0.89401899
## motor_premotor_SMA.lh.thickness 2.80489444 -0.65068283
## perisylvian.rh.thickness -0.94393787 1.15847060
## ventralfrontal.rh.thickness -0.27982984 -0.45174958
## ventromedialoccipital.lh.thickness -2.39680348 0.37093793
## inferiorparietal.rh.thickness -0.20881417 0.99827580
## ventromedialoccipital.rh.thickness -0.86594446 -0.15440560
## superiorparietal.rh.thickness 0.06444381 1.01733021
## medialtemporal.lh.thickness 1.07207894 -0.72919504
## dorsolateralprefrontal.lh.area 0.90469439 -0.65364085
## motorpremotor.lh.area 0.88366726 -0.79375138
## superiorparietal.lh.thickness 0.57336220 -0.93346662
## dorsolateralprefrontal.rh.area -0.45090801 0.68606122
## occipital.rh.area 0.57617753 -0.31328688
## occipital.rh.thickness 0.57615386 -0.67486217
## motorpremotor.rh.area -0.98407355 -0.24873178
## ventralfrontal.lh.thickness -0.35130305 -0.03719715
## precuneus.rh.area 0.54329017 -0.82107902
## occipital.lh.thickness 0.25107936 -1.01417277
## medialtemporal.rh.thickness 0.24565262 0.06611091
## parsopercularis.lh.area -0.92713325 0.49447561
## parsopercularis.rh.area -0.31262077 -0.39751903
## precuneus.lh.area 0.56040020 -0.42510178
## superiortemporal.lh.area -1.14151848 0.17263804
## pval.TD_vs_Good pval.TD_vs_Poor
## middletemporal.rh.thickness 0.005525174 0.123169627
## posterolateraltemporal.rh.area 0.717461549 0.031502420
## anteromedialtemporal.lh.area 0.705717336 0.061143805
## superiorparietal.rh.area 0.524457792 0.131365824
## medialprefrontal.lh.thickness 0.677931016 0.076823579
## middletemporal.lh.thickness 0.510365327 0.055839716
## superiorparietal.lh.area 0.501128867 0.134218641
## dorsomedialfrontal.lh.area 0.105303557 0.678914597
## medialprefrontal.rh.thickness 0.520219365 0.272092547
## temporalpole.rh.thickness 0.182515299 0.180307013
## orbitofrontal.rh.area 0.004061395 0.004699057
## temporalpole.lh.thickness 0.955185137 0.082178529
## occipital.lh.area 0.085711306 0.599117724
## posterolateraltemporal.lh.area 0.625172512 0.344914593
## inferiorparietal.lh.area 0.360719517 0.053174065
## dorsolateralprefrotal.lh.thickness 0.695928937 0.179219372
## inferiorparietal.lh.thickness 0.302548979 0.238756489
## orbitofrontal.lh.area 0.356621001 0.017053979
## perisylvian.lh.thickness 0.345928755 0.387474182
## superiortemporal.rh.area 0.276033472 0.742270900
## dorsolateralprefrotal.rh.thickness 0.530791893 0.881541209
## anteromedialtemporal.rh.area 0.584134133 0.389823804
## dorsomedialfrontal.rh.area 0.795932285 0.036403210
## inferiorparietal.rh.area 0.027775608 0.867560414
## motor_premotor_SMA.rh.thickness 0.132216658 0.028039496
## motor_premotor_SMA.lh.thickness 0.036291546 0.006299476
## perisylvian.rh.thickness 0.799240172 0.348008513
## ventralfrontal.rh.thickness 0.373565377 0.780321412
## ventromedialoccipital.lh.thickness 0.188691907 0.018841738
## inferiorparietal.rh.thickness 0.131472705 0.835117540
## ventromedialoccipital.rh.thickness 0.222317948 0.389078318
## superiorparietal.rh.thickness 0.266544921 0.948775623
## medialtemporal.lh.thickness 0.826342416 0.286868405
## dorsolateralprefrontal.lh.area 0.502991170 0.368310330
## motorpremotor.lh.area 0.837802144 0.379491172
## superiorparietal.lh.thickness 0.775320195 0.567987958
## dorsolateralprefrontal.rh.area 0.922916239 0.653260634
## occipital.rh.area 0.574293259 0.566092762
## occipital.rh.thickness 0.960400139 0.566108683
## motorpremotor.rh.area 0.347765765 0.328009702
## ventralfrontal.lh.thickness 0.413944279 0.726273163
## precuneus.rh.area 0.910871155 0.588422089
## occipital.lh.thickness 0.710808117 0.802388604
## medialtemporal.rh.thickness 0.976117097 0.806572522
## parsopercularis.lh.area 0.765585637 0.356611722
## parsopercularis.rh.area 0.829277289 0.755372241
## precuneus.lh.area 0.939040433 0.576753254
## superiortemporal.lh.area 0.946182305 0.257017722
## pval.Good_vs_Poor fdrq.TD_vs_Good
## middletemporal.rh.thickness 0.206535221 0.1326042
## posterolateraltemporal.rh.area 0.008483109 0.9761171
## anteromedialtemporal.lh.area 0.008594918 0.9761171
## superiorparietal.rh.area 0.058690208 0.9761171
## medialprefrontal.lh.thickness 0.233139903 0.9761171
## middletemporal.lh.thickness 0.176733527 0.9761171
## superiorparietal.lh.area 0.069929008 0.9761171
## dorsomedialfrontal.lh.area 0.030364342 0.7933000
## medialprefrontal.rh.thickness 0.489926021 0.9761171
## temporalpole.rh.thickness 0.640132452 0.9057212
## orbitofrontal.rh.area 0.797139090 0.1326042
## temporalpole.lh.thickness 0.053516450 0.9761171
## occipital.lh.area 0.393097427 0.7933000
## posterolateraltemporal.lh.area 0.346835702 0.9761171
## inferiorparietal.lh.area 0.288387881 0.9437441
## dorsolateralprefrotal.lh.thickness 0.554542121 0.9761171
## inferiorparietal.lh.thickness 0.068871440 0.9437441
## orbitofrontal.lh.area 0.094907165 0.9437441
## perisylvian.lh.thickness 0.853664547 0.9437441
## superiortemporal.rh.area 0.802876807 0.9437441
## dorsolateralprefrotal.rh.thickness 0.341467101 0.9761171
## anteromedialtemporal.rh.area 0.624206920 0.9761171
## dorsomedialfrontal.rh.area 0.171710503 0.9761171
## inferiorparietal.rh.area 0.218216386 0.4354985
## motor_premotor_SMA.rh.thickness 0.374290497 0.7933000
## motor_premotor_SMA.lh.thickness 0.517322378 0.4354985
## perisylvian.rh.thickness 0.250500151 0.9761171
## ventralfrontal.rh.thickness 0.652807328 0.9437441
## ventromedialoccipital.lh.thickness 0.711772955 0.9057212
## inferiorparietal.rh.thickness 0.321488880 0.7933000
## ventromedialoccipital.rh.thickness 0.877721941 0.9437441
## superiorparietal.rh.thickness 0.312402770 0.9437441
## medialtemporal.lh.thickness 0.468247853 0.9761171
## dorsolateralprefrontal.lh.area 0.515425366 0.9761171
## motorpremotor.lh.area 0.429948432 0.9761171
## superiorparietal.lh.thickness 0.353699051 0.9761171
## dorsolateralprefrontal.rh.area 0.494877863 0.9761171
## occipital.rh.area 0.754968477 0.9761171
## occipital.rh.thickness 0.501924742 0.9761171
## motorpremotor.rh.area 0.804276395 0.9437441
## ventralfrontal.lh.thickness 0.970430749 0.9761171
## precuneus.rh.area 0.414311663 0.9761171
## occipital.lh.thickness 0.313896344 0.9761171
## medialtemporal.rh.thickness 0.947472757 0.9761171
## parsopercularis.lh.area 0.622475737 0.9761171
## parsopercularis.rh.area 0.692160839 0.9761171
## precuneus.lh.area 0.672030293 0.9761171
## superiortemporal.lh.area 0.863420323 0.9761171
## fdrq.TD_vs_Poor fdrq.Good_vs_Poor
## middletemporal.rh.thickness 0.4294997 0.8010153
## posterolateraltemporal.rh.area 0.2496220 0.2062780
## anteromedialtemporal.lh.area 0.2934903 0.2062780
## superiorparietal.rh.area 0.4294997 0.4795132
## medialprefrontal.lh.thickness 0.3287141 0.8010153
## middletemporal.lh.thickness 0.2934903 0.8010153
## superiorparietal.lh.area 0.4294997 0.4795132
## dorsomedialfrontal.lh.area 0.8575763 0.4795132
## medialprefrontal.rh.thickness 0.6237181 0.8010153
## temporalpole.rh.thickness 0.5091022 0.8704098
## orbitofrontal.rh.area 0.1511874 0.8977969
## temporalpole.lh.thickness 0.3287141 0.4795132
## occipital.lh.area 0.7988236 0.8010153
## posterolateraltemporal.lh.area 0.6237181 0.8010153
## inferiorparietal.lh.area 0.2934903 0.8010153
## dorsolateralprefrotal.lh.thickness 0.5091022 0.8318132
## inferiorparietal.lh.thickness 0.6237181 0.4795132
## orbitofrontal.lh.area 0.2261009 0.5694430
## perisylvian.lh.thickness 0.6237181 0.9158838
## superiortemporal.rh.area 0.8798973 0.8977969
## dorsolateralprefrotal.rh.thickness 0.9002974 0.8010153
## anteromedialtemporal.rh.area 0.6237181 0.8704098
## dorsomedialfrontal.rh.area 0.2496220 0.8010153
## inferiorparietal.rh.area 0.9002974 0.8010153
## motor_premotor_SMA.rh.thickness 0.2496220 0.8010153
## motor_premotor_SMA.lh.thickness 0.1511874 0.8010153
## perisylvian.rh.thickness 0.6237181 0.8010153
## ventralfrontal.rh.thickness 0.8798973 0.8704098
## ventromedialoccipital.lh.thickness 0.2261009 0.8760283
## inferiorparietal.rh.thickness 0.8907920 0.8010153
## ventromedialoccipital.rh.thickness 0.6237181 0.9158838
## superiorparietal.rh.thickness 0.9487756 0.8010153
## medialtemporal.lh.thickness 0.6237181 0.8010153
## dorsolateralprefrontal.lh.area 0.6237181 0.8010153
## motorpremotor.lh.area 0.6237181 0.8010153
## superiorparietal.lh.thickness 0.7988236 0.8010153
## dorsolateralprefrontal.rh.area 0.8474733 0.8010153
## occipital.rh.area 0.7988236 0.8977969
## occipital.rh.thickness 0.7988236 0.8010153
## motorpremotor.rh.area 0.6237181 0.8977969
## ventralfrontal.lh.thickness 0.8798973 0.9704307
## precuneus.rh.area 0.7988236 0.8010153
## occipital.lh.thickness 0.8798973 0.8010153
## medialtemporal.rh.thickness 0.8798973 0.9676318
## parsopercularis.lh.area 0.6237181 0.8704098
## parsopercularis.rh.area 0.8798973 0.8743084
## precuneus.lh.area 0.7988236 0.8718231
## superiortemporal.lh.area 0.6237181 0.9158838
Make effect size maps
dotSize = 6
fontSize=28
fdr_thresh = 0.05
# output_res_fdr = output_res[output_res$fdrq<=fdr_thresh,]
# output_res_fdr[order(output_res_fdr$pval),]
#------------------------------------------------------------------------------
# df2plot = output_res[4:nrow(output_res),]
df2plot = output_res
df2plot$region = rownames(df2plot)
new_region_labels = c("motor-premotor","occipital","posterolateral temporal","superior parietal",
"orbitofrontal","superior temporal","inferior parietal","dorsomedial frontal",
"anteromedial temporal","precuneus","dorsolateral prefrontal","pars opercularis,subcentral",
"motor-premotor","occipital","posterolateral temporal","superior parietal",
"orbitofrontal","superior temporal","inferior parietal","dorsomedial frontal",
"anteromedial temporal","precuneus","dorsolateral prefrontal","pars opercularis,subcentral",
"motor-premotor-supplementary motor area","ventral frontal",
"ventromedial occipital","inferior parietal","middle temporal","perisylvian",
"dorsolateral prefrontal","occipital","temporal pole","superior parietal",
"medial prefrontal","medial temporal",
"motor-premotor-supplementary motor area","ventral frontal",
"ventromedial occipital","inferior parietal","middle temporal","perisylvian",
"dorsolateral prefrontal","occipital","temporal pole","superior parietal",
"medial prefrontal","medial temporal")
hemi_labels = c("LH","LH","LH","LH",
"LH","LH","LH","LH",
"LH","LH","LH","LH",
"RH","RH","RH","RH",
"RH","RH","RH","RH",
"RH","RH","RH","RH",
"LH","LH","LH","LH",
"LH","LH","LH","LH",
"LH","LH","LH","LH",
"RH","RH","RH","RH",
"RH","RH","RH","RH",
"RH","RH","RH","RH")
metric_labels = c("SA","SA","SA","SA",
"SA","SA","SA","SA",
"SA","SA","SA","SA",
"SA","SA","SA","SA",
"SA","SA","SA","SA",
"SA","SA","SA","SA",
"CT","CT","CT","CT",
"CT","CT","CT","CT",
"CT","CT","CT","CT",
"CT","CT","CT","CT",
"CT","CT","CT","CT",
"CT","CT","CT","CT")
df2plot$region = new_region_labels
df2plot$hemisphere = hemi_labels
df2plot$metric = metric_labels
colors2use = get_ggColorHue(7)
#------------------------------------------------------------------------------
# Surface Area TD vs ASD Good Effect size maps
# LH
# df2use = subset(d2plot, d2plot$hemisphere=="LH")
df2use = subset(df2plot, df2plot$hemisphere=="LH" & df2plot$metric=="SA")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pl = ggseg(.data = df2use,
atlas = chenAr,
mapping=aes(fill=es.TD_vs_Good),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="left") +
# scale_fill_manual(values = colors2use[1:3]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
# RH
# df2use = subset(d2plot, d2plot$hemisphere=="RH")
df2use = subset(df2plot, df2plot$hemisphere=="RH" & df2plot$metric=="SA")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pr = ggseg(.data = df2use,
atlas = chenAr,
mapping=aes(fill=es.TD_vs_Good),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="right") +
# scale_fill_manual(values = colors2use[4:7]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
pf = (pr)/(pl)
ggsave(filename = file.path(plotdir, "SA_TD_vs_ASDGood_effectSizeMaps.pdf"), plot = pf)
pf
#------------------------------------------------------------------------------
# Surface Area TD vs ASD Poor Effect size maps
# LH
# df2use = subset(d2plot, d2plot$hemisphere=="LH")
df2use = subset(df2plot, df2plot$hemisphere=="LH" & df2plot$metric=="SA")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pl = ggseg(.data = df2use,
atlas = chenAr,
mapping=aes(fill=es.TD_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="left") +
# scale_fill_manual(values = colors2use[1:3]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
# RH
# df2use = subset(d2plot, d2plot$hemisphere=="RH")
df2use = subset(df2plot, df2plot$hemisphere=="RH" & df2plot$metric=="SA")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pr = ggseg(.data = df2use,
atlas = chenAr,
mapping=aes(fill=es.TD_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="right") +
# scale_fill_manual(values = colors2use[4:7]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
pf = (pr)/(pl)
ggsave(filename = file.path(plotdir, "SA_TD_vs_ASDPoor_effectSizeMaps.pdf"), plot = pf)
pf
#------------------------------------------------------------------------------
# Surface Area ASD Good vs ASD Poor Effect size maps
# LH
# df2use = subset(d2plot, d2plot$hemisphere=="LH")
df2use = subset(df2plot, df2plot$hemisphere=="LH" & df2plot$metric=="SA")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pl = ggseg(.data = df2use,
atlas = chenAr,
mapping=aes(fill=es.Good_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="left") +
# scale_fill_manual(values = colors2use[1:3]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
# RH
# df2use = subset(d2plot, d2plot$hemisphere=="RH")
df2use = subset(df2plot, df2plot$hemisphere=="RH" & df2plot$metric=="SA")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pr = ggseg(.data = df2use,
atlas = chenAr,
mapping=aes(fill=es.Good_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="right") +
# scale_fill_manual(values = colors2use[4:7]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
pf = (pr)/(pl)
ggsave(filename = file.path(plotdir, "SA_ASDGood_vs_ASDPoor_effectSizeMaps.pdf"), plot = pf)
pf
#------------------------------------------------------------------------------
# Cortical Thickness TD vs ASD Good Effect size maps
# LH
# df2use = subset(d2plot, d2plot$hemisphere=="LH")
df2use = subset(df2plot, df2plot$hemisphere=="LH" & df2plot$metric=="CT")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pl = ggseg(.data = df2use,
atlas = chenTh,
mapping=aes(fill=es.TD_vs_Good),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="left") +
# scale_fill_manual(values = colors2use[1:3]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
# RH
# df2use = subset(d2plot, d2plot$hemisphere=="RH")
df2use = subset(df2plot, df2plot$hemisphere=="RH" & df2plot$metric=="CT")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pr = ggseg(.data = df2use,
atlas = chenTh,
mapping=aes(fill=es.TD_vs_Good),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="right") +
# scale_fill_manual(values = colors2use[4:7]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
pf = (pr)/(pl)
ggsave(filename = file.path(plotdir, "CT_TD_vs_ASDGood_effectSizeMaps.pdf"), plot = pf)
pf
#------------------------------------------------------------------------------
# Cortical Thickness TD vs ASD Poor Effect size maps
# LH
# df2use = subset(d2plot, d2plot$hemisphere=="LH")
df2use = subset(df2plot, df2plot$hemisphere=="LH" & df2plot$metric=="CT")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pl = ggseg(.data = df2use,
atlas = chenTh,
mapping=aes(fill=es.TD_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="left") +
# scale_fill_manual(values = colors2use[1:3]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
# RH
# df2use = subset(d2plot, d2plot$hemisphere=="RH")
df2use = subset(df2plot, df2plot$hemisphere=="RH" & df2plot$metric=="CT")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pr = ggseg(.data = df2use,
atlas = chenTh,
mapping=aes(fill=es.TD_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="right") +
# scale_fill_manual(values = colors2use[4:7]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
pf = (pr)/(pl)
ggsave(filename = file.path(plotdir, "CT_TD_vs_ASDPoor_effectSizeMaps.pdf"), plot = pf)
pf
#------------------------------------------------------------------------------
# Cortical Thickness ASD Good vs ASD Poor Effect size maps
# LH
# df2use = subset(d2plot, d2plot$hemisphere=="LH")
df2use = subset(df2plot, df2plot$hemisphere=="LH" & df2plot$metric=="CT")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pl = ggseg(.data = df2use,
atlas = chenTh,
mapping=aes(fill=es.Good_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="left") +
# scale_fill_manual(values = colors2use[1:3]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
# RH
# df2use = subset(d2plot, d2plot$hemisphere=="RH")
df2use = subset(df2plot, df2plot$hemisphere=="RH" & df2plot$metric=="CT")
df2use$label2fill = factor(c(1:dim(df2use)[1]))
pr = ggseg(.data = df2use,
atlas = chenTh,
mapping=aes(fill=es.Good_vs_Poor),
# mapping=aes(fill=Fstat),
# mapping=aes(fill=label2fill),
position="stacked",
colour="black",
hemisphere="right") +
# scale_fill_manual(values = colors2use[4:7]) + guides(fill=FALSE) +
# scale_fill_gradientn(colors = colorRampPalette(c("white","red"))(100), limits = c(0,15)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100), limits = c(-0.7,0.7)) +
# guides(fill = guide_colourbar(nbin = 100)) +
guides(fill=guide_legend(title="Effect Size")) +
theme(text = element_text(size=fontSize))
pf = (pr)/(pl)
ggsave(filename = file.path(plotdir, "CT_ASDGood_vs_ASDPoor_effectSizeMaps.pdf"), plot = pf)
pf
Make A-P and D-V and Effect Size Gradient Plots
regions2use = colnames(fs_data)[5:(ncol(fs_data)-4)]
full_regnames = c("motor-premotor","occipital","posterolateral temporal","superior parietal",
"orbitofrontal","superior temporal","inferior parietal","dorsomedial frontal",
"anteromedial temporal","precuneus","dorsolateral prefrontal","pars opercularis,subcentral",
"motor-premotor","occipital","posterolateral temporal","superior parietal",
"orbitofrontal","superior temporal","inferior parietal","dorsomedial frontal",
"anteromedial temporal","precuneus","dorsolateral prefrontal","pars opercularis,subcentral",
"motor-premotor-supplementary motor area","ventral frontal","ventromedial occipital","inferior parietal",
"middle temporal","perisylvian","dorsolateral prefrontal","occipital",
"temporal pole","superior parietal","medial prefrontal","medial temporal",
"motor-premotor-supplementary motor area","ventral frontal","ventromedial occipital","inferior parietal",
"middle temporal","perisylvian","dorsolateral prefrontal","occipital",
"temporal pole","superior parietal","medial prefrontal","medial temporal")
sa_ordering = c(1,12,7,10,
4,6,9,3,
8,11,2,5)
ct_ordering = c(1,7,6,3,
10,4,11,5,
8,2,12,9)
gensim_order = c(sa_ordering,sa_ordering, ct_ordering,ct_ordering)
df4plot = output_res[regions2use,c("es.TD_vs_Good","es.TD_vs_Poor","es.Good_vs_Poor")]
df4plot$phenotype = "CT"
df4plot$phenotype[1:24] = "SA"
df4plot$phenotype = factor(df4plot$phenotype)
df4plot$hemisphere = "RH"
df4plot$hemisphere[1:12] = "LH"
df4plot$hemisphere[25:36] = "LH"
df4plot$hemisphere = factor(df4plot$hemisphere)
df4plot$full_region_names = factor(full_regnames)
df4plot$gensim_gradient = gensim_order
ct_df = subset(df4plot, df4plot$phenotype=="CT")
sa_df = subset(df4plot, df4plot$phenotype=="SA")
ct_df$cluster = "dorsal"
ct_df$cluster[c(6,7,8,9,10,11,12)] = "ventral"
ct_df$cluster = factor(ct_df$cluster)
sa_df$cluster = "anterior"
sa_df$cluster[c(6,7,8,9,10,11,12)] = "posterior"
sa_df$cluster = factor(sa_df$cluster)
dotSize = 10
fontSize = 18
p = ggplot(data = ct_df, aes(x = gensim_gradient, y = es.TD_vs_Good, fill = es.TD_vs_Good)) +
geom_point(size=dotSize, colour="black", pch=21) +
geom_smooth(method=lm, colour="black") + xlim(0,12) +
scale_x_continuous(breaks=seq(1,12,1)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100)) +
guides(colour=FALSE, fill=FALSE) +
xlab("Brain Region") +
ylab("Effect Size (Cohen's d)") +
ggtitle("CT TD vs ASD Good") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"geneSimGradient_TD_vs_ASDGood_CT_effectSize.pdf"))
p
cor.test(ct_df$gensim_gradient, ct_df$es.TD_vs_Good)
##
## Pearson's product-moment correlation
##
## data: ct_df$gensim_gradient and ct_df$es.TD_vs_Good
## t = -1.5876, df = 22, p-value = 0.1267
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.6410888 0.0950901
## sample estimates:
## cor
## -0.3206047
p = ggplot(data = ct_df, aes(x = cluster, y = es.TD_vs_Good, fill=cluster))
p = p + geom_jitter(size=dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
# scale_fill_manual(values = c("#1e90ff","#ff8d1e")) +
scale_fill_manual(values = c("blue","red")) +
guides(colour=FALSE, fill=FALSE) +
xlab("Clusters") +
ylab("Effect Size (Cohen's d)") +
ggtitle("Dorsal-Ventral Clusters TD vs ASD Good") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"dvGradient_TD_vs_ASDGood_CT_effectSize.pdf"))
p
t.test(es.TD_vs_Good ~ cluster, data = ct_df)
##
## Welch Two Sample t-test
##
## data: es.TD_vs_Good by cluster
## t = 0.026067, df = 21.915, p-value = 0.9794
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.1712679 0.1756271
## sample estimates:
## mean in group dorsal mean in group ventral
## -0.007408816 -0.009588433
p = ggplot(data = ct_df, aes(x = gensim_gradient, y = es.TD_vs_Poor, fill = es.TD_vs_Poor)) +
geom_point(size=dotSize, colour="black", pch=21) +
geom_smooth(method=lm, colour="black") + xlim(0,12) +
scale_x_continuous(breaks=seq(1,12,1)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100)) +
guides(colour=FALSE, fill=FALSE) +
xlab("Brain Region") +
ylab("Effect Size (Cohen's d)") +
ggtitle("CT TD vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"geneSimGradient_TD_vs_ASDPoor_CT_effectSize.pdf"))
p
cor.test(ct_df$gensim_gradient, ct_df$es.TD_vs_Poor)
##
## Pearson's product-moment correlation
##
## data: ct_df$gensim_gradient and ct_df$es.TD_vs_Poor
## t = -0.79935, df = 22, p-value = 0.4326
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.5351309 0.2525090
## sample estimates:
## cor
## -0.1680004
p = ggplot(data = ct_df, aes(x = cluster, y = es.TD_vs_Poor, fill=cluster))
p = p + geom_jitter(size=dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
# scale_fill_manual(values = c("#1e90ff","#ff8d1e")) +
scale_fill_manual(values = c("blue","red")) +
guides(colour=FALSE, fill=FALSE) +
xlab("Clusters") +
ylab("Effect Size (Cohen's d)") +
ggtitle("Dorsal-Ventral Clusters TD vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"dvGradient_TD_vs_ASDPoor_CT_effectSize.pdf"))
p
t.test(es.TD_vs_Poor ~ cluster, data = ct_df)
##
## Welch Two Sample t-test
##
## data: es.TD_vs_Poor by cluster
## t = -1.0499, df = 12.781, p-value = 0.3132
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.4389992 0.1521861
## sample estimates:
## mean in group dorsal mean in group ventral
## -0.05882051 0.08458602
p = ggplot(data = ct_df, aes(x = gensim_gradient, y = es.Good_vs_Poor, fill = es.Good_vs_Poor)) +
geom_point(size=dotSize, colour="black", pch=21) +
geom_smooth(method=lm, colour="black") + xlim(0,12) +
scale_x_continuous(breaks=seq(1,12,1)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100)) +
guides(colour=FALSE, fill=FALSE) +
xlab("Brain Region") +
ylab("Effect Size (Cohen's d)") +
ggtitle("CT ASD Good vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"geneSimGradient_ASDGood_vs_ASDPoor_CT_effectSize.pdf"))
p
cor.test(ct_df$gensim_gradient, ct_df$es.Good_vs_Poor)
##
## Pearson's product-moment correlation
##
## data: ct_df$gensim_gradient and ct_df$es.Good_vs_Poor
## t = 0.66563, df = 22, p-value = 0.5126
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.2786875 0.5147273
## sample estimates:
## cor
## 0.140505
p = ggplot(data = ct_df, aes(x = cluster, y = es.Good_vs_Poor, fill=cluster))
p = p + geom_jitter(size=dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
# scale_fill_manual(values = c("#1e90ff","#ff8d1e")) +
scale_fill_manual(values = c("blue","red")) +
guides(colour=FALSE, fill=FALSE) +
xlab("Clusters") +
ylab("Effect Size (Cohen's d)") +
ggtitle("Dorsal-Ventral Clusters ASD Good vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"dvGradient_ASDGood_vs_ASDPoor_CT_effectSize.pdf"))
p
t.test(es.Good_vs_Poor ~ cluster, data = ct_df)
##
## Welch Two Sample t-test
##
## data: es.Good_vs_Poor by cluster
## t = -1.2707, df = 10.307, p-value = 0.2318
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.36781016 0.09995858
## sample estimates:
## mean in group dorsal mean in group ventral
## -0.04447174 0.08945405
p = ggplot(data = sa_df, aes(x = gensim_gradient, y = es.TD_vs_Good, fill = es.TD_vs_Good)) +
geom_point(size=dotSize, colour="black", pch=21) +
geom_smooth(method=lm, colour="black") + xlim(0,12) +
scale_x_continuous(breaks=seq(1,12,1)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100)) +
guides(colour=FALSE, fill=FALSE) +
xlab("Brain Region") +
ylab("Effect Size (Cohen's d)") +
ggtitle("SA TD vs ASD Good") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"geneSimGradient_TD_vs_ASDGood_SA_effectSize.pdf"))
p
cor.test(sa_df$gensim_gradient, sa_df$es.TD_vs_Good)
##
## Pearson's product-moment correlation
##
## data: sa_df$gensim_gradient and sa_df$es.TD_vs_Good
## t = 0.054893, df = 22, p-value = 0.9567
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.3935522 0.4131487
## sample estimates:
## cor
## 0.01170234
p = ggplot(data = sa_df, aes(x = cluster, y = es.TD_vs_Good, fill=cluster))
p = p + geom_jitter(size=dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
# scale_fill_manual(values = c("#1e90ff","#ff8d1e")) +
scale_fill_manual(values = c("blue","red")) +
guides(colour=FALSE, fill=FALSE) +
xlab("Clusters") +
ylab("Effect Size (Cohen's d)") +
ggtitle("Anterior-Posterior Clusters TD vs ASD Good") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"apGradient_TD_vs_ASDGood_SA_effectSize.pdf"))
p
t.test(es.TD_vs_Good ~ cluster, data = sa_df)
##
## Welch Two Sample t-test
##
## data: es.TD_vs_Good by cluster
## t = 1.1522, df = 16.828, p-value = 0.2654
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.08611438 0.29297353
## sample estimates:
## mean in group anterior mean in group posterior
## 0.03023708 -0.07319250
p = ggplot(data = sa_df, aes(x = gensim_gradient, y = es.TD_vs_Poor, fill = es.TD_vs_Poor)) +
geom_point(size=dotSize, colour="black", pch=21) +
geom_smooth(method=lm, colour="black") + xlim(0,12) +
scale_x_continuous(breaks=seq(1,12,1)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100)) +
guides(colour=FALSE, fill=FALSE) +
xlab("Brain Region") +
ylab("Effect Size (Cohen's d)") +
ggtitle("SA TD vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"geneSimGradient_TD_vs_ASDPoor_SA_effectSize.pdf"))
p
cor.test(sa_df$gensim_gradient, sa_df$es.TD_vs_Poor)
##
## Pearson's product-moment correlation
##
## data: sa_df$gensim_gradient and sa_df$es.TD_vs_Poor
## t = -0.57123, df = 22, p-value = 0.5736
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.4999111 0.2969860
## sample estimates:
## cor
## -0.1208943
p = ggplot(data = sa_df, aes(x = cluster, y = es.TD_vs_Poor, fill=cluster))
p = p + geom_jitter(size=dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
# scale_fill_manual(values = c("#1e90ff","#ff8d1e")) +
scale_fill_manual(values = c("blue","red")) +
guides(colour=FALSE, fill=FALSE) +
xlab("Clusters") +
ylab("Effect Size (Cohen's d)") +
ggtitle("Anterior-Posterior Clusters TD vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"apGradient_TD_vs_ASDPoor_SA_effectSize.pdf"))
p
t.test(es.TD_vs_Poor ~ cluster, data = sa_df)
##
## Welch Two Sample t-test
##
## data: es.TD_vs_Poor by cluster
## t = 1.6048, df = 11.985, p-value = 0.1345
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.07913237 0.52140838
## sample estimates:
## mean in group anterior mean in group posterior
## 0.09699741 -0.12414060
p = ggplot(data = sa_df, aes(x = gensim_gradient, y = es.Good_vs_Poor, fill = es.Good_vs_Poor)) +
geom_point(size=dotSize, colour="black", pch=21) +
geom_smooth(method=lm, colour="black") + xlim(0,12) +
scale_x_continuous(breaks=seq(1,12,1)) +
scale_fill_gradientn(colors = colorRampPalette(c("blue","white","red"))(100)) +
guides(colour=FALSE, fill=FALSE) +
xlab("Brain Region") +
ylab("Effect Size (Cohen's d)") +
ggtitle("SA ASD Good vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"geneSimGradient_ASDGood_vs_ASDPoor_SA_effectSize.pdf"))
p
cor.test(sa_df$gensim_gradient, sa_df$es.Good_vs_Poor)
##
## Pearson's product-moment correlation
##
## data: sa_df$gensim_gradient and sa_df$es.Good_vs_Poor
## t = -0.58651, df = 22, p-value = 0.5635
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## -0.5023311 0.2940367
## sample estimates:
## cor
## -0.1240773
p = ggplot(data = sa_df, aes(x = cluster, y = es.Good_vs_Poor, fill=cluster))
p = p + geom_jitter(size=dotSize, colour="black", pch=21) +
geom_boxplot(fill = NA, colour = "#000000", outlier.shape = NA) +
# scale_fill_manual(values = c("#1e90ff","#ff8d1e")) +
scale_fill_manual(values = c("blue","red")) +
guides(colour=FALSE, fill=FALSE) +
xlab("Clusters") +
ylab("Effect Size (Cohen's d)") +
ggtitle("Anterior-Posterior Clusters ASD Good vs ASD Poor") +
theme(text = element_text(size=fontSize),
axis.text.x = element_text(size=fontSize),
axis.text.y = element_text(size=fontSize),
plot.title = element_text(size=fontSize,hjust=0.5))
ggsave(filename = file.path(plotdir,"apGradient_ASDGood_vs_ASDPoor_SA_effectSize.pdf"))
p
t.test(es.Good_vs_Poor ~ cluster, data = sa_df)
##
## Welch Two Sample t-test
##
## data: es.Good_vs_Poor by cluster
## t = 0.71299, df = 9.2556, p-value = 0.4934
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.2262792 0.4358494
## sample estimates:
## mean in group anterior mean in group posterior
## 0.06145044 -0.04333464